Dimerization-Induced Corepressor Binding and Relaxed DNA-Binding Specificity Are Critical for PML RARA-Induced Immortalization

Home , Corepressor

Dimerization-induced corepressor binding and relaxed DNA-binding specificity are critical for PML͞RARA-induced immortalization

Jun Zhou†‡§, Laurent Pe´re`s†, Nicole Honore´ †, Rihab Nasr†, Jun Zhu†‡¶, and Hugues de The´ †‡¶

†Centre National de la Recherche Scientifique Unite´Mixte de Recherche 7151, Universite´de Paris 7, Equipe labelliseé par la Ligue Nationale contre le Cancer, Hoˆpital St. Louis, 1 Avenue C. Vellefaux, 75475 Paris Cedex 10, France, ‡Poˆle de Recherche Franco–Chinois en Sciences du Vivant et de Geńomique and §Shanghai Institute of Hematology, Rui-Jin Hospital, 197 Rui-Jin Road II, Shanghai 200025, People’s Republic of China

Communicated by Zhu Chen, Shanghai Second Medical University, Shanghai, People’s Republic of China, April 27, 2006 (received for review March 15, 2006) The pathogenesis of acute promyelocytic leukemia involves the fusions [PML and promyelocytic leukemia zinc finger (PLZF)] transcriptional repression of master genes of myeloid differentia- not only provide a dimerization interface and an additional tion by the promyelocytic leukemia–retinoic acid receptor ␣ (PML͞ repression domain to the fusion (2–4, 7, 17), but they could also RARA) oncogene. PML-enforced RARA homodimerization allows contribute to transformation through deregulation of the path- the tighter binding of corepressors, silencing RARA target genes. In ways normally controlled by PML or PLZF (1). By transducing addition, homodimerization dramatically extends the spectrum of a set of RARA mutants in primary hematopoietic progenitor DNA-binding sites of the fusion protein compared with those cells, we establish that dimerization-induced enhanced SMRT of normal RARA. Yet, any contribution of these two properties of binding and repression of PML͞RARA-specific targets are both PML͞RARA to differentiation arrest and immortalization of pri- critical to differentiation arrest and immortalization, demon- mary mouse hematopoietic progenitors was unknown. We dem- strating how dimerization converts RARA into an oncogenic onstrate that dimerization-induced silencing mediator of retinoid protein. and thyroid receptors (SMRT)-enhanced binding and relaxed DNA- binding site specificity are both required for efficient immortaliza- Results tion. Thus, enforced RARA dimerization is critical not only for RARA Homodimerization Is Required for Transformation of Primary triggering transcriptional repression but also for extending the Mouse Hematopoietic Progenitors. Primary hematopoietic pro- repertoire of target genes. Our studies exemplify how dimeriza- genitor cells undergo a sharp differentiation arrest and be- tion-induced gain of functions converts an unessential transcrip- come immortal after transduction of PML͞RARA (7, 18, 19). tion factor into a dominant oncogenic protein. Self-dimerizing RARA mutants, such as p50-RARA, which recruit SMRT and efficiently silence nuclear receptor target chromatin ͉ leukemia ͉ retinoic acid ͉ transcription genes, were proposed to play a critical role in APL pathogenesis (5, 6). Yet, this fusion failed to immortalize mouse he t(15;17) chromosomal translocation, specific for acute progenitors (Fig. 1), consistent with the requirement of an Tpromyelocytic leukemia (APL), fuses the promyelocytic additional small ubiquitin-like modifier of protein (SUMO)- leukemia (PML) gene to the retinoic acid receptor ␣ (RARA) dependent repression for transformation of primary cells (7). gene. APL leukemogenesis primarily involves transcriptional In the context of PML͞RARA, the transcriptional corepressor silencing of critical differentiation genes by the PML-RARA Daxx can replace this SUMO-dependent repression domain fusion protein (1), which results in both a differentiation block (7). Expression of a Daxx-RARA fusion protein also failed to and expansion of myeloid progenitors. Yet, although several immortalize or induce a significant differentiation block in structure͞function studies have been performed in various im- these progenitor cells (Fig. 1). mortalized cell lines, with transcriptional repression or differ- To address the contribution of RARA dimers to transfor- entiation block as an outcome, the molecular determinants of mation, we inserted the tetramerization domain of p53 (not differentiation arrest and͞or immortalization in primary hema- including the sumoylation site) into Daxx-RARA. When trans- topoietic cells are largely unexplored. At the molecular level, two duced into progenitor cells, Daxx-tet-RARA was as efficient distinct domains of the fusion protein cooperate to enforce as PML͞RARA in blocking their differentiation and immor- transcriptional repression: (i) the RARA hormone-binding do- talizing them (Fig. 1), thus demonstrating the crucial role of mains, dimerized by the PML coiled coil, recruit silencing dimerization in transformation of primary cells ex vivo.To mediator of retinoid and thyroid receptors (SMRT) with a much assess directly the role of enhanced SMRT recruitment in greater affinity than RARA͞retinoid X receptor (RXR) het- immortalization, mutations in RARA previously shown to erodimers (2–6); (ii) a specific sumoylation site in PML is abrogate SMRT binding (4) were introduced into Daxx-tet- required for full repression (7). In contrast to RAR͞RXR RARA. Although these mutations impaired immortalization, heterodimers, which bind DNA sites that consist of two AG- GTCA core motifs in a direct-repeat orientation, with spacing of 1, 2, or 5 (8, 9), PML͞RARA forms homodimers or multimers Conflict of interest statement: No conflicts declared. with RXR that tightly bind two core motifs, in any orientation, Freely available online through the PNAS open access option. even with very wide spacing (10–12). Accordingly, PML͞RARA Abbreviations: APL, acute promyelocytic leukemia; DR, direct repeat; Luc, luciferase; MEF, was first shown to silence RA target genes (13), then type II mouse embryo fibroblast; NCoR, nuclear receptor corepressor; PLZF, promyelocytic leuke- nuclear receptor targets (10, 11), and finally de novo target genes mia zinc finger; PML, promyelocytic leukemia protein; RA, retinoic acid; RARA, retinoic acid ͞ receptor ␣; RXR, retinoid X receptor; SMRT, silencing mediator of retinoid and thyroid such as type II transglutaminase or CCAAT enhancer-binding receptors; SUMO, small ubiquitin-like modifier of protein; tet, tetramerization domain; Tk, protein ␤, many of which contain widely spaced binding sites (12, thymidine kinase. 14–16). Whether these de novo targets of the fusion protein are ¶To whom correspondence may be addressed. E-mail: [email protected] or simple bystanders or play a critical role in transformation is [email protected]. unknown. The upstream partners of RARA in the X-RARA © 2006 by The National Academy of Sciences of the USA

9238–9243 ͉ PNAS ͉ June 13, 2006 ͉ vol. 103 ͉ no. 24 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0603324103 Downloaded by guest on September 30, 2021 Fig. 1. Comparison of primary hematopoietic precursors transduced by RARA mutants. (A) Flow cytometry analysis of c-Kit, Gr1, and Mac1 expression in RARA mutant-transduced cells. tet, tetramerization; NCoR, nuclear receptor corepressor. (B) Serial methyl cellulose replating of transduced cells analyzed in A. The means of four different experiments are shown. (C) May–Grunwald–Giemsa-stained preparations of cells analyzed in A.

demonstrating the importance of enhanced SMRT binding to not shown). Thus, Daxx-tet-RARA is similar to PML͞RARA this process, Daxx-tet-RARA*NCoR expression nevertheless with respect to SMRT and DNA binding as well as transforma- triggered a very significant differentiation block when ex- tion ex vivo. pressed in myeloid progenitors (Fig. 1C). Comparable levels of these different RARA fusions were expressed after retroviral Control of Gene Expression by RARA Mutants. To establish that the transduction (Fig. 2A). As expected, only PML͞RARA was different RARA fusions described above regulate distinct sets of able to delocalize PML nuclear bodies (data not shown), but genes, as expected from the different binding site specificity, we all of the RARA fusions had a finely trabecular and mi- first expressed them in immortalized mouse fibroblasts, where all crospeckled nuclear localization (data not shown). Altogether, three RARs were excised by Cre recombinase (20). Thus, in these findings demonstrate that Daxx and SMRT repression these cells, any retinoic acid (RA)-induced change in gene domains, as well as dimerization, are indispensable for com- transcription is exclusively dependent on a transduced RARA- plete differentiation block and immortalization of primary derived receptor. We first transiently transfected DR5- or DR12- hematopoietic precursor cells ex vivo. containing reporter plasmids (12) and examined the effect of RA on luciferase activity. Remarkably, whereas RARA or Daxx- Binding Specificity of RARA Mutants. To demonstrate directly that RARA activated the DR5 reporter almost exclusively, PML͞ Daxx-tet-RARA shares with PML͞RARA altered binding site RARA, Daxx-tet-RARA, or Daxx-tet-RARA*NCoR activated specificity, we monitored the binding of these proteins onto either of the DR5 or DR12 reporters after the addition of direct-response elements DR5, DR8, and DR12. A specific hormone (Fig. 3A). retarded protein–DNA complex was observed when extracts of To demonstrate further the relaxed specificity of target-gene Daxx-tet-RARA-transfected CHO cells, but not untransfected activation by RARA homodimers, we analyzed RA-dependent ones, were added to either of the probes, demonstrating that transcriptional activation of three target genes by quantitative Daxx-tet-RARA, like PML͞RARA (12), can bind an extended PCR. We focused on three genes: two well characterized RARA repertoire of response elements (Fig. 2 and data not shown). targets (RARB and CYP26A1) and a primary PML͞RARA Identical results were obtained in the presence of RXR (data not target, transglutaminase type II (14). As expected, for the RARB shown). or CYP26A1 genes, RA activated all of the RARA-derived Enforced RARA dimerization results in an increased affinity receptors (Fig. 3A), the strongest activation being observed for for SMRT (5, 6). When bacterially expressed SMRT was added RARA and Daxx-tet-RARA*NCoR. Enhanced binding of to binding reactions, PML͞RARA or Daxx-tet-RARA bound corepressors onto self-dimerizing RARA fusions blunted the SMRT very avidly, whereas, as expected, RARA-RXR, Daxx- response to RA, as expected. Remarkably, proteins containing RARA, or Daxx-tet-RARA*NCoR did not (Fig. 2B and data a dimerization domain (notably, PML͞RARA, Daxx-tet- MEDICAL SCIENCES

Zhou et al. PNAS ͉ June 13, 2006 ͉ vol. 103 ͉ no. 24 ͉ 9239 Downloaded by guest on September 30, 2021 Fig. 2. Expression and binding properties of RARA mutants. (A) Western blot analysis of transduced primary hematopoietic cells (antibody: RARA). (B) Gel-shift analysis of RARA-RXR, PML-RARA, Daxx-tet-RARA, and Daxx-tet-RARA*NCoR in response to the consensus DR5, DR8, or DR12 response element in the presence or absence of bacterially expressed SMRT. Self-dimerizing fusions yield two complexes (dimers and multimers), as shown in ref. 12. Note that the RARA͞RXR complex binds with decreasing afﬁnity to DR5, DR8, and DR12, whereas fusion proteins containing a dimerization domain bind the three types of probe with similar afﬁnity, as shown for PML͞RARA (12).

RARA, and Daxx-tet-RARA*NCoR) all activated the trans- induce a differentiation block and did not immortalize primary glutaminase type II gene, whereas Daxx-RARA and wild-type hematopoietic progenitors (Fig. 4A). RARA failed to do so. This observation strongly suggests that To demonstrate unambiguously that both of the two repres- this gene contains an as yet unidentified response element sion domains are functional in this large fusion protein, we specific for RARA dimers, and it points to the importance compared the ability of Daxx-RARA and Daxx-SMRT-RARA of RARA dimerization in the extension of the target-gene to regulate transcription in the different settings explored above. repertoire. In contrast to Daxx-RARA, Daxx-SMRT-RARA was incapable of activating the endogenous RARB or CYP26A1 gene in trans- Altered Binding Site Specificity by Dimerization Is Key to the APL duced RAR-null cells (Fig. 4B). In transient transfections by a ϩ/ϩ Phenotype. The findings presented above demonstrate that Daxx- DR5-derived reporter (8) of RAR MEFs stably transduced tet-RARA recapitulates all of the transformation- and tran- by Daxx-SMRT-RARA, this fusion blunted the RA response mediated by endogenous RARs, whereas Daxx-RARA en- scription-associated phenotypes of PML͞RARA, presumably hanced luciferase induction by RA (Fig. 4C). Similar results were through dimerization-induced SMRT recruitment and͞or re- obtained in HeLa or COS cells transiently cotransfected with the laxed DNA-binding site specificity. Defective immortalization by reporter plasmid and an expression vector for Daxx-SMRT- Daxx-tet-RARA*NCoR demonstrates that corepressor binding RARA (data not shown). Finally, in RARϩ/ϩ MEFs, activation is essential for transformation of primary cells by RARA of the endogenous RARB gene by RA was sharply diminished by homodimers. To investigate the importance of dimerization- stably expressed Daxx-SMRT-RARA, but it was activated by induced extension in the target-gene repertoire, we generated Daxx-RARA (Fig. 4C). That Daxx-SMRT-RARA does not Daxx-SMRT-RARA, a construct in which the repression do- significantly affect the differentiation of hematopoietic progen- main of SMRT (21) was inserted into Daxx-RARA, thus disso- itors cells, despite the efficient tethering of the two critical ciating from RARA homodimerization the enhanced binding of repression domains onto RAR͞RXR targets, demonstrates that SMRT onto RARA. Importantly, this construct (which was altered DNA sequence recognition elicited by enforced RARA expressed at levels similar to the other RARA fusions) did not homodimerization is also critical to the APL phenotype.

9240 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0603324103 Zhou et al. Downloaded by guest on September 30, 2021 Fig. 3. Transcriptional regulation by RARA mutants. (A) Induction of luciferase activity by RA (10Ϫ6 M, 12 h) in RARϪ/Ϫ mouse embryo ﬁbroblasts (MEFs) stably transduced by the indicated RARA fusion proteins. DR5- and DR12-containing reporter plasmids are indicated. One representative experiment of three is shown. (B–D) RARϪ/Ϫ MEFs stably expressing the indicated RARA mutants were treated or not with 1 ␮M all-trans-RA for 12 h, and CYP26A1, RARB, transglutaminase type II (TGII) gene expression was analyzed by quantitative PCR, as indicated.

Discussion demonstration of immortalization requires more than three RARA dimerization triggers both the efficient recruitment of passages in methyl cellulose (7) and that the FKBP artificial SMRT-NCoR corepressors (5, 6) and a dramatically relaxed dimerization module contains a consensus sumoylation site (26) DNA-binding specificity (12). In the context of APL fusion that may complicate the interpretation of experiments per- proteins, PLZF and PML also provide an additional repression formed with FKBP-enforced RARA dimers (22). domain (2–4, 7). These three events cooperate to repress any Relaxed DNA-binding specificity by RARA homodimers also gene that contains two AGGTCA half-sites in any orientation or appears to be critical to the immortalization of primary cells, spacing (12, 15). Although previous studies in cell lines had strongly suggesting that non-RARA targets of the fusions play demonstrated the importance of RARA dimerization and a critical role in transformation and identifying a critical gain of SMRT binding in the differentiation block (4–6), the contribu- function of PML͞RARA. This conclusion is in line with the tion of these three gains of function of APL fusions in primary observation that a dimerization-deficient PLZF͞RARA mutant cells was not established. failed to transform myeloid progenitors, although it retained a This work, consistent with three very recently published greatly enhanced binding to the SMRT corepressor at the reports (22–24), demonstrates the critical role of RARA dimer- biochemical level (22). Note that the v-ErbA oncogene (27) and ization in the immortalization of primary myeloid progenitors. PLZF͞RARA (D. Kamashev and H.d.T., unpublished observa- Note that all APL-associated RARA fusion proteins contain a tions) also have an extended repertoire of DNA-binding sites potent dimerization domain (25). Although RARA dimeriza- compared with their parental nuclear receptors, suggesting that tion through the rapamycin-sensitive FKBP domain was sug- dimerization-induced change in DNA-binding specificity may be gested to suffice for ex vivo transformation (22), our own results shared by other oncogenic transcription factors. with p50-RARA (Fig. 1), as previously with PML͞RARA- Our experiments in primary cells also confirm the critical K160R (7), suggest that enforced RARA dimerization is insuf- importance of corepressor binding previously shown in the ficient per se. Although this discordance may relate to the U937-differentiation assay (4). Paradoxically, in primary cells, a experimental conditions, such as the mouse strains used or the very significant differentiation block remains with the Daxx-tet- cytokine content of the assay, we note that, in our hands, RARA*NCoR fusion, whereas immortalization is abolished MEDICAL SCIENCES

Zhou et al. PNAS ͉ June 13, 2006 ͉ vol. 103 ͉ no. 24 ͉ 9241 Downloaded by guest on September 30, 2021 Fig. 4. Relaxed binding site speciﬁcity is essential for PML-RARA transformation. (A) Immunophenotyping, replating assays, morphology, and Western blot analyses of Daxx-SMRT-RARA-transduced primary hematopoietic progenitors. (B) Quantitative PCR analysis of target genes in RARϪ͞Ϫ MEFs transduced by Daxx-SMRT-RARA or Daxx-RARA. TGII, transglutaminase type II. (C Left) Induction of luciferase activity by RA (10Ϫ6 M, 12 h) in RARϩ/ϩ MEFs transduced by Daxx-SMRT-RARA or Daxx-RARA (DR5-Tk-Luc reporter). Tk, thymidine kinase. (C Right) RARB gene expression analyzed by quantitative PCR in the same cells. A representative experiment is shown.

(Fig. 1). Conversely, in PML͞RARA, SUMO-dependent repres- mice and culture of the transduced progenitors cells with G418 sion was required for the full differentiation block ex vivo or in selection in methyl cellulose with stem cell factor, IL-3, IL-6, and vivo, despite the efficient recruitment of SMRT (7). Finally, granulocyte͞macrophage colony-stimulating factor were per- massive RARA overexpression induces a major differentiation formed as described in ref. 18. After a week, neomycin-selected block, but it does not immortalize primary myeloid progenitor cells were recovered from methyl cellulose and either analyzed cells in our hands (data not shown). This finding strongly (by FACS, May–Grunwald–Giemsa staining, immunofluores- suggests that the molecular determinants of the differentiation cence, and Western blotting) or replated at a density of 10,000 block and the immortalization by RARA fusions are not iden- cells per dish. Cells were serially replated until they stopped tical. growing. Anti-RARA rabbit serum (RP115) was used for im- The experiments reported here demonstrate that enforced munofluorescence and Western blotting. Daxx-RARA was de- RARA homodimerization triggers at least two critical gains of scribed in ref. 28. The tetramerization domain of human p53 function (enhanced repression and extended target-gene reper- (amino acids 324–355) was inserted into a Daxx-RARA con- toire) that cooperate to yield differentiation arrest and immor- struct, yielding Daxx-tet-RARA. The repression domain of talization. That a RARA fusion (Daxx-tet-RARA) devoid of any PML sequence can fully recapitulate PML͞RARA-induced mouse SMRT (amino acids 1–1031) was inserted into a Daxx- RARA construct, yielding Daxx-SMRT-RARA. The corepres- transformation could suggest that the only contribution of PML ͞ is to provide the SUMO-dependent repression domain and sor-binding sites in RARA (A194T H195P) were mutated in the homodimerization, indicating that disruption of PML nuclear Daxx-tet-RARA construct Daxx-tet-RARA*NCoR, as de- bodies and PML signaling might be dispensable in this ex vivo scribed in ref. 4. All of these constructs were cloned in a pMSCV setting. The complex issue of the minimal requirements for retroviral vector, and the virus was transiently produced by efficient RARA-induced leukemogenesis should now be ad- transfection of Plat-E cells (29). dressed in vivo. Analysis of the Properties of Transduced Proteins. Electrophoretic Materials and Methods mobility-shift analyses were performed as described in ref. 12 by Retroviral Transduction and Cell Analyses. Infection of lineage- using extracts from CHO transfected cells and a bacterially depleted bone marrow from 5-fluorouracil-treated C57BL͞6 expressed SMRT fragment (3). CHO cells transiently trans-

9242 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0603324103 Zhou et al. Downloaded by guest on September 30, 2021 fected with pSG5-RARA mutants were lysed in ice for 30 min Luc (8, 12)] were transfected into RARϩ/ϩ or RARϪ/Ϫ MEFs in the lysis buffer [50 mM Tris, pH 8.0͞0.6% Igepal CA-630 transduced with the RARA-derived fusions, and luciferase (Sigma–Aldrich)͞480 mM NaCl͞10% (vol/vol) glycerol͞0.1 mM activity was monitored as previously described. EDTA͞1mMDTT͞1.5 mM sodium vanadate͞0.4 mM PMSF͞3 ␮g/ml aprotinin͞1 ␮g/ml pepstatin͞1 ␮g/ml leupeptin]. The Note. While this manuscript was being reviewed, it was shown that centrifugation supernatant was stored at Ϫ80°C. PCR- dominant-negative RARA or fusion of histone deacetylase to RARA is constructed DNA probes were gel-purified and labeled as de- unable to initiate APL development in transgenic mice, fully consistent scribed above and mixed with cellular extract in the binding with the data reported here (30). The AML1-ETO fusion was also shown buffer as above with 0.02 ␮g͞␮l poly(dI-dC) at room tempera- to undergo multimerization, and multimerization with the expected ture. DNA⅐protein complexes and free DNA were separated by changes in DNA-binding specificity was shown to be required for electrophoresis on a 3.75% polyacrylamide gel. transformation (31). We thank C. Lavau and V. Lallemand-Breitenbach for helpful discus- Cell Culture and Transient Transfection. Immortalized MEFs in sions, F. Brau and N. Settlerbad for confocal imaging, and M. T. Daniel which the three RARs were excised (20) were retrovirally trans- and A. Janin for help with slide interpretation. We thank H. Gronemeyer duced with the different mutants, grown in DMEM containing 10% ␣ ␤ ␥ ␮ for floxed RAR ,- , and - MEFs. SMRT and p50-RARA were kind FBS, and then treated with 1 M RA (Sigma–Aldrich) overnight. gifts of R. Evans. This work was supported by the Actions The´matiques Expressions of RARB, CYP26A1, and transglutaminase type II et Incitatives sur Programme, the Programme de Recherche en Re´seau genes were monitored with the Light Cycler system and gene- Franco–Chinois, and the Alliance des Recherches sur le Cancer. J. Zhou specific TaqMan probes (Applied Biosystems). The quantification was supported by the French Ministry for Foreign Affairs, the Centre data were analyzed with LightCycler and its software (Roche). National de la Recherche Scientifique, and the Association for Virus ␤2-Microglobulin was used as an internal control. Cancer Prevention; R.N. was supported by the Eli Lilly International Reporter plasmids [RARE-3-Tk-Luc and DR5- (or 12)-Tk- Foundation.

1. Melnick, A. & Licht, J. D. (1999) Blood 93, 3167–3215. 16. Duprez, E., Wagner, K., Koch, H. & Tenen, D. G. (2003) EMBO J. 22, 2. Lin, R. J., Nagy, L., Inoue, S., Shao, W. L., Miller, W. H. & Evans, R. M. (1998) 5806–5816. Nature 391, 811–814. 17. Dong, S., Zhu, J., Reid, A., Strutt, P., Guidez, F., Zhong, H. J., Wang, Z. Y., 3. He, L.-Z., Guidez, F., Tribioli, C., Peruzzi, D., Ruthardt, M., Zelent, A. & Licht, J., Waxman, S., Chomienne, C., et al. (1996) Proc. Natl. Acad. Sci. USA Pandolfi, P. P. (1998) Nat. Genet. 18, 126–135. 93, 3624–3629. 4. Grignani, F., de Matteis, S., Nervi, C., Tomassoni, L., Gelmetti, V., Cioce, M., 18. Du, C., Redner, R. L., Cooke, M. P. & Lavau, C. (1999) Blood 94, 793–802. Fanelli, M., Ruthardt, M., Ferrara, F. F., Zamir, I., et al. (1998) Nature 391, 19. Insinga, A., Monestiroli, S., Ronzoni, S., Carbone, R., Pearson, M., Pruneri, G., 815–818. Viale, G., Appella, E., Pelicci, P. & Minucci, S. (2004) EMBO J. 23, 1144–1154. 5. Lin, R. & Evans, R. (2000) Mol. Cell 5, 821–830. 20. Altucci, L., Rossin, A., Hirsch, O., Nebbioso, A., Vitoux, D., Wilhelm, E., 6. Minucci, S., Maccarana, M., Cioce, M., De Luca, P., Gelmetti, V., Segalla, S., Guidez, F., Schiavone, E. M., Grimwade, D., Zelent, A., et al. (2005) Cancer Di Croce, L., Giavara, S., Matteucci, C., Gobbi, A., et al. (2000) Mol. Cell 5, Res. 65, 8754–8765. 811–820. 21. Ordentlich, P., Downes, M., Xie, W., Genin, A., Spinner, N. B. & Evans, R. M. 7. Zhu, J., Zhou, J., Peres, L., Riaucoux, F., Honore´, N., Kogan, S. & de The´, H. (1999) Proc. Natl. Acad. Sci. USA 96, 2639–2644. (2005) Cancer Cell 7, 143–153. 22. Kwok, C., Zeisig, B. B., Dong, S. & So, C. W. (2006) Cancer Cell 9, 95–108. 8. de The´,H., Vivanco-Ruiz, M. M., Tiollais, P., Stunnenberg, H. & Dejean, A. 23. Sternsdorf, T., Phan, V. T., Maunakea, M. L., Ocampo, C., Sohal, J., Siletto, (1990) Nature 343, 177–180. A., Galimi, F., Le Beau, M. M., Evans, R. & Kogan, S. (2006) Cancer Cell 9, 9. Perlmann, T., Rangarajan, P. N., Umesono, K. & Evans, R. M. (1993) Genes 81–94. Dev. 7, 1411–1422. 24. Licht, J. D. (2006) Cancer Cell 9, 73–74. 10. Perez, A., Kastner, P., Sethi, S., Lutz, Y., Reibel, C. & Chambon, P. (1993) 25. Piazza, F., Gurrieri, C. & Pandolfi, P. P. (2001) Oncogene 20, 7216–7222. EMBO J. 12, 3171–3182. 26. Rodriguez, M. S., Dargemont, C. & Hay, R. T. (2001) J. Biol. Chem. 276, 11. Jansen, J. H., Mahfoudi, A., Rambaud, S., Lavau, C., Wahli, W. & Dejean, A. 12654–12659. (1995) Proc. Natl. Acad. Sci. USA 92, 7401–7405. 27. Zubkova, I. & Subauste, J. S. (2004) Mol. Biol. Rep. 31, 131–137. 12. Kamashev, D. E., Vitoux, D. & de The´, H. (2004) J. Exp. Med. 199, 1–13. 28. Friedman, P. N., Chen, X., Bargonetti, J. & Prives, C. (1993) Proc. Natl. Acad. 13. de The´,H., Lavau, C., Marchio, A., Chomienne, C., Degos, L. & Dejean, A. Sci. USA 90, 3319–3323, and correction (1993) 90, 5878. (1991) Cell 66, 675–684. 29. Morita, S., Kojima, T. & Kitamura, T. (2000) Gene Ther. 12, 1063–1066. 14. Benedetti, L., Grignani, F., Scicchitano, B. M., Jetten, A. M., Diverio, D., Lo 30. Matsushita, H., Scaglioni, P. P., Bhaumik, M., Rego, E. M., Cai, L. F., Majid, Coco, F., Avvisati, G., Gambarcoti-Passerini, C., Adamo, S., Levin, A. A., et al. S. M., Miyachi, H., Kakizuka, A., Miller, W. H., Jr., & Pandolfi, P. P. (2006) (1996) Blood 87, 1939–1950. J. Exp. Med. 203, 821–828. 15. Meani, N., Minardi, S., Licciulli, S., Gelmetti, V., Coco, F. L., Nervi, C., Pelicci, 31. Liu, Y., Cheney, M. D., Gaudet, J. J., Chruszcz, M., Lukasik, S. M., Sugiyama, P. G., Muller, H. & Alcalay, M. (2005) Oncogene 24, 3358–3368. D., Lary, J., Cole, J., Dauter, Z., Minor, W., et al. (2006) Cancer Cell 9, 249–260. MEDICAL SCIENCES

Zhou et al. PNAS ͉ June 13, 2006 ͉ vol. 103 ͉ no. 24 ͉ 9243 Downloaded by guest on September 30, 2021